...I agree that this would be a useful test. But it would be a useful test only if it is agreed beforehand to accept well known lexicographers' opinions as to whether words are semantically similar or distinct. While certainly some of these opinions are debatable, it is unlikely that they have been deliberately chosen to support or counter the hypothesis being tested, whereas the personal opinions of any of us involved in this debate may be framed with the hypothesis in mind. I would propose following the opinions of the compilers of the Westminster Hebrew morphological database, which is widely available in computer Bible packages.
In our numerically driven world, that observation can either be
verified or falsified by doing a statistical study, but no one has
done it. So at the moment it remains an observation, neither proven
nor falsified.
Archive powered by MHonArc 2.6.24.